INDEX

Explanations

variance, average, complexity, performance

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

 Marca

1.32

 marched

1.29

霧

1.27

 Sand

1.26

 cabeza

1.26

 Salam

1.25

 Câmara

1.24

 Anak

1.22

 Boul

1.21

 Bark

1.20

POSITIVE LOGITS

 innate

1.11

 kiến

1.08

ablen

1.01

omorphisms

1.00

 evolutionary

0.98

 methionine

0.96

invariant

0.95

 incompar

0.95

苷

0.95

 optimal

0.94

Activations Density 0.190%